22 Prof. E. Wilson .
[ May 9 , distribution is reached in the rigid phase , and the new phase , as Prof. Trouton has shown , must be considered as bivariant till the film has attained a certain limiting thickness .
It would be interesting to speculate as to the subsequent course of Prof. Trouton 's curves , and I hope that his further researches will throw light on the subject .
] Effects of Self-induction in Iron Cylinder* By Professor Ernest Wilson , King 's College , London .
( Communicated by Sir William H. Preece , K.C.B. , F.R.S. Received May 9 , \#151 ; Read June 7 , 1906 .
) If a solid cylindrical conductor be divided into imaginary concentric tubular conductors , the ordinary self and mutual induction theory shows that when the conductor is subjected to an alternating potential difference , the interior shells carry electric currents of smaller density than the exterior ones , and the currents suffer a phase displacement greater the nearer the centre of the cylinder .
The theory shows that the permeability and conductivity of the material play an important part , the effects above mentioned being increased in magnitude with increasing permeability and conductivity .
When , however , the material of the conductor has variable permeability the problem becomes more complicated , and it is the object of this paper to examine more closely what goes on in an iron cylinder when electric currents are reversed in it , and maintained steady after reversal .
A second part of this research will deal with alternating currents of varying frequency and wave-form .
The cylinder employed is of mild steel and has a diameter and length each equal to 10 inches ( 25'4 cm .
) .
It is provided with holes drilled in a plane containing its axis of figure in such a manner that exploring coils can be threaded to enclose certain portions of that plane .
The exploring coils are three in number .
They are each 2 inches wide in a direction parallel with the axis of figure and midway between the ends of the Cylinder .
Their depths in a radial direction are 1 , 2 , and 2 inches , and their average radii are 0'5 , 2 and 4 inches respectively .
These coils are referred to as Coils Nos. 1 , 2 , and 3 , No. 1 being near the centre of the cylinder .
The cylinder has been already described , f but for the purpose of * In connection with this research , I wish to acknowledge a grant voted to me by the Council of the Royal Society- out of the- Government Grant Fund .
t ' Roy .
Soc. Proc. , ' vol. 69 , p. 440 .
1906 .
] Effects of Self-induction in an Iron Cylinder .
the present research has had a hole ^-inch diameter drilled through it coinciding with the axis of figure .
At each end of the cylinder are two massive gunmetal projections , which in the present research serve to conduct the electric current into the cylinder .
The conductors attached to these projections are connected to a reversing switch so constructed that at its mid position it short-circuits the circuit of the cylinder and its conductors , which were arranged in the form of a circle about 6 feet diameter .
The electric current was supplied by storage cells through an adjustable resistance and the shunt of an ampere-meter .
The potential difference of the cells and the adjustable resistance were such that on reversing the current in the cylinder its value in the main circuit remained constant .
In fact , the reversal of the main current was practically instantaneous .
The epoch of reversal was noted on a seconds clock and at two-second intervals after reversal the deflections of the dead-beat galvanometers in circuit with the exploring coils were noted .
The deflections have been reduced to volts per turn per square centimetre of the coils from which they were obtained , and plotted in terms of the time .
Figs. 1 , 2 , and 3 give the results obtained from Coils Nos. 1 , 2 , and 3 respectively , and each curve is numbered to correspond with the total amperes reversed when it was observed .
The curves have been integrated in order to find the maximum average value of the induction density B for the respective coils .
The average values are set out in Table I. Current Density .
\#171 ; .
If the current density over the cross-section of the cylinder was constant , the force H would vary as the radius .
If the values of B for Coil No. 3 in Table I be plotted in terms of the total currents reversed in the cylinder , it will be found that the resulting curve resembles the BH curve of a piece of mild steel .
If the values of B for Coils Nos. 2 and 1 be plotted in terms of the total current multiplied respectively by 0*5 and 0T25 ( to correspond with their radii ) , it will be seen that , although similar , the three curves do not lie on one another .
To make them do so the coefficients with which to multiply the total currents are 0*75 and 0'28 respectively .
This suggests that under steady conditions the current density is greater near the centre of the cylinder than near the surface .
The average relative densities appear to be : ( 1 ) over the area of the cylinder within the average radius of Coil No. 1 , 0-56 ; ( 2 ) over the annulus between the average radii of Coils Nos. 1 and 2 , 0'36 ; ( 3 ) over the annulus between the average radii of Coils Nos. 2 and 3 , 0'21 .
A total current of 950 amperes corresponds to an average force H of about 6 C.G-.S .
units for Coil No. 3 , and the BH curve for mild steel is well Prof. E. Wilson .
[ May 9 , represented by the results of the experiments .
The relation between the average value of H for each of the coils and the total current is given in Table I. Table I. Total current in cylinder .
Coil No. 1 .
Ay .
H. 1-68 1*18 0-924 0-75 0-627 0-426 0-372 0-283 0-218 0-141 0-079 Max. ay .
B. Coil No. 2 .
Ay .
H. 4-5 3T5 2-48 2-01 1-69 1 -14 0-998 0-758 0-583 0-379 0-213 Max. av .
B. Coil No. 3 .
Ay .
H. .68 .25 .52 .33 .01 0-777 0-505 0-284 Max. ay .
B. Comparison of E.M.F. Curves .
* The E.M.F. curves are given in figs. 1 , 2 , 3 , to which reference will be made .
Coil No. 3 experiences its maximum rate of change at once , although after reversal of about 400 amperes there is slight evidence of a second maximum at about 20 seconds after reversal .
After about 400 amperes it will be seen that as the total current increases the curves cross one another at shorter intervals , indicating that the effects penetrate more rapidly after the maximum average induction density has passed the value at which maximum permeability occurs .
\#163 ; o O ' CQ ec \#163 ; .
F , '\amp ; \#166 ; 1 Is C* ^ \ ^8 70 Co il 1 .
\#163 ; \#163 ; S / 240 7/ 80 \l ^634 ^621 ^360 0 10 20 30 4 0 SO00 70 80 90 100 110 120 TIME AFTER REVERSAL .
IN SECONDS .
1906 .
] Effects of Self-induction in an Iron Cylinder .
Coil No. 2 shows the effect of a second maximum more markedly .
When the total current in the cylinder is small there is no second maximum .
At about 200 amperes the second maximum shows signs of being developed , and with larger currents its development is such as to make it the most important feature of the curves .
As in the case of Coil No. 3 , the E.M.F. curves cross one another at earlier intervals after about 420 amperes has been reached .
TINE AFTER REVERSAL , IN SECONDS .
Coil No. 1 shows similar effects , but its E.M.F. 's have a still later maximum for a given current .
Moreover , the time between the first and second maxima is greater than in the case of Coil No. 2 .
These results are in keeping with what has already been observed in the case of the reversal of currents in a copper coil surrounding cylinders of 4 and 12 inches diameter , * and in the present cylinder when rotated in a magnetic field.'f It is difficult to say exactly how long the effects take to die away owing to the ultimate want of sensibility of * 'Phil .
Trans. , ' A , vol. 186 ( 1895 ) , pp. 93\#151 ; 121 ; also 'Journ .
Inst. Elec .
Eng. , ' Part 116 , vol. 24 , p. 194 .
t ' Roy .
Soc. Proc. , ' vol. 69 , p. 435 , and vol. 70 , p. 359 .
Prof. E. Wilson .
[ May 9 , Coil 3 .
10 20 30 TIME AFTER REVERSAL IN SECONDS .
the instruments , but a comparison of the times taken to practically die away is of interest .
In Table II are given the times taken to reverse the magnetism at the centre of 4- and 12-inch diameter cylinders when the currents in their magnetising coils are instantaneously reversed , and then maintained steady .
The times in these two cases roughly vary as the square of the diameters of the cylinders .
The results obtained in the present experiments are also included in Table II .
Table II .
4-inch magnet .
12-inch magnet .
10-inch cylinder .
Duration Duration Duration in seconds .
Total in Max. H. in Max. H. \#151 ; current in seconds .
seconds .
Coil No. 1 .
Coil No. 2 .
Coil No. 3 .
cylinder .
40 1 -7 360 1 -2 20 45 45 3 0 420 2 -4 20 28 15 80 33 4-96 180 6-0 30 48 30 160 10 16 -0 80 11 *0 58 80 50 210 5 37 -0 50 24 -0 75 100 \#151 ; 240 116 112 60 356 120 120 50 424 190 90 , \#151 ; 522 90 75 40 670 90 65 30 950 1906 .
] Effects of Self-induction an Iron Cylinder .
Upon integrating the E.M.F. curves it was found that the average magnetic flux , for total currents up to about 240 amperes , was reversed in sign for Ooil No. 2 after Coil No. 1 .
For currents greater than 240 amperes the % curves cross the axis of time in the order 3 , 2 , 1 .
In all cases , however , the total interior average currents , as obtained from the magnetic hysteresis loops , reversed in the order 3 , 2 , 1 .
Application of Results to other Sections .
Comparing two cylinders whose diameters are as 1 : the value of H at similar radii will be the same when the total currents are as 1 : n. Considering unit length of the two cylinders , the total magnetic induction up to similar radii varies as n , and the electric resistance of ' similarly placed longitudinal paths varies as 1 / n2 .
Therefore to induce n times the current in those paths the E.M.F. 's must vary as If the time varies as 2 the E.M.F. 's will vary as 1/ n , thereby giving rise to the same value of the magnetic force H at similar radii .
A paper recently published* dealt with the self-induction of bull-headed , railway rails , weighing 70 lbs. per yard .
It was there shown how greatly the self-induction varies with the frequency for a given current , and with the current for a given frequency .
The head of one of these rails is roughly equivalent to a cylinder of 2 inches diameter .
A current of 100 amperes in such a cylinder corresponds to 500 amperes in our 10-inch cylinder if the forces at similar radii are to be the same .
We infer roughly that an alternating current of 8 seconds periodic time and approximately rectilinear wave-form would permit of practically the whole section being made use of as regards conduction for a small fraction of the time of each half period .
The frequencies employed in practice are of the order 25 periods per second , and are enormous by comparison .
In railway work a _FL-shaped rail would obviously be more suited from the standpoint of electric conduction of alternating currents .
A current of about 0'5 ampere in an iron wire of 0T inch diameter would give rise to a force of about 0'3 near the surface .
Changes of magnetism in our 10-inch cylinder were observed 20 seconds after reversal of about 50 amperes .
A frequency of 250 would allow of the full section of the wire being made use of during a small fraction of the time of each half period with 0*5 ampere , but not with a current of 5 amperes .
In conclusion , I wish to thank Mr. A. E. O'Dell for his patience and care in working out results , and Mr. H. W. Franks for his assistance in the experimental part of the paper .
* ' The Electrician , ' February 23 , 1906 .

